- Email: [email protected]
Distribution and stability of aflatoxin M1 during processing and storage of Minas Frescal cheese
Food Control 24 (2012) 104e108
Contents lists available at SciVerse ScienceDirect
Food Control journal homepage: www.elsevier.com/locate/foodcont
Distribution and stability of aflatoxin M1 during processing and storage of Minas Frescal cheese A.M. Fernandesa, B. Corrêab, R.E. Rosimc, E. Kobashigawac, C.A.F. Oliveirac, * a
Department of Basic Sciences, School of Animal Science and Food Engineering, University of São Paulo, Av. Duque de Caxias Norte, 225, CEP 13635-900, Pirassununga, SP, Brazil Department of Microbiology, Institute of Biomedical Sciences, University of São Paulo, Av. Prof. Lineu Prestes, 1374, CEP 05508-900, São Paulo, SP, Brazil c Department of Food Engineering, School of Animal Science and Food Engineering, University of São Paulo, Av. Duque de Caxias Norte, 225, CEP 13635-900, Pirassununga, SP, Brazil b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 May 2011 Received in revised form 12 August 2011 Accepted 6 September 2011
Aflatoxin M1 (AFM1) is a hepatocarcinogen found in milk of animals that have consumed feeds with aflatoxin B1. The carry-over of AFM1 from milk to Minas Frescal cheese produced with or without starter cultures was determined. 40 L of milk were divided into 10 L each and assigned to the following treatments for cheese manufacture: 0.250 ng AFM1 mL1, 0.500 ng AFM1 mL1, 0.250 ng AFM1 mL1 þ starter, 0.500 ng AFM1 mL1 þ starter. Quantification of AFM1 was achieved by high performance liquid chromatography. The carry-over of AFM1 from milk to cheese ranged from 30.64% to 42.26%. There was no effect of storage time on AFM1. Milk with AFM1 in levels studied may concentrate the toxin in Minas Frescal cheese, but at concentrations below the Brazilian tolerance limit. The addition of starter cultures did not influence concentration or stability of the AFM1 in cheese over 30 days storage. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Mycotoxins AFM1 Cheese Residues
1. Introduction Aflatoxin M1 (AFM1) is a hepatocarcinogenic metabolite found in milk of animals that have consumed feeds contaminated with aflatoxin B1, AFB1 (Park & Pohland, 1986), the major mycotoxin produced by fungi of the genus Aspergillus (Moss, 1998). The carryover of AFM1 into milk has been demonstrated in several studies (Wood, 1991) and the rate of transfer of AFB1 from feed to milk in dairy cows is between 1 and 3% with a mean transfer of 1.7% (Jouany & Diaz, 2005). The occurrence of AFM1 in milk and dairy products is a public health concern since the International Agency for Research on Cancer (1993) has classified AFM1 as a probable human carcinogen (Group 2). There are several reports of AFM1 contamination of cheeses produced from milk artificially treated with AFB1; however, these reports refer to major cheeses from each country and, according to López, Ramos, Ramadan, Bulacio, and Perez (2001), the AFM1 levels in cheese seem to be variable, depending on the type of the product. Studies involving different types of cheeses demonstrated the presence of higher levels of AFM1 in cheese than found in the artificially contaminated milk (Elgerbi, Aidoo, Candlish, & Tester, 2004; Oruc, Cibik, Yilmaz, & Kalkanli, 2006). The distribution of
* Corresponding author. Tel.: þ55 19 3565 4173; fax: þ55 19 3565 4284. E-mail address: [email protected] (C.A.F. Oliveira). 0956-7135/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.foodcont.2011.09.010
AFM1 during cheese manufacture also varies with the type of final product. Fresh cheese produced in Argentina from artificially contaminated milk with AFM1 at levels of 1.7e2.0 ng mL1 had 60% of AFM1 in the whey and 40% in cheese (López et al., 2001). In Brazil, there is little information on the occurrence of AFM1 in cheeses manufactured in the country, except for two reports, one of them by Prado et al. (2000), who observed that 74.7% of Minas Frescal samples from the state of Minas Gerais showed AFM1 at levels ranging from 0.02 to 6.92 ng g1. In a recent study, Oliveira, Franco, Rosim, and Fernandes (2011) found a high incidence of AFM1 in Minas Frescal cheeses produced in the state of São Paulo, but levels were usually below the tolerance limit (2.5 ng g1) established by Brazilian regulations (Agência Nacional de Vigilância Sanitária, 2011). Other countries have established different tolerance limits for AFM1 in cheeses: 0.20 ng g1 in the Netherlands, and 0.25 ng g1 in Austria, Switzerland and Turkey (Food and Agriculture Organization, 2004). Several studies have suggested that starter culture bacteria used in cheese manufacture might inactivate AFM1 by forming a linkage between AFM and lactic bacteria (Pierides, El-Nizami, Peltonen, Salminen, & Ahokas, 2000). In addition, AFM1 may form linkages with milk proteins such as casein depending on the influence of acidity (Brackett & Marth, 1982). Minas Frescal cheese is typical Brazilian cheese classified as a semi-fat, non-ripened and high humidity cheese which is obtained by enzymatic coagulation of milk with rennet and with or without starter cultures. Although in
A.M. Fernandes et al. / Food Control 24 (2012) 104e108
Brazil previous studies indicated a high incidence of AFM1 in fluid milk, there are no previous reports on the carry-over of AFM1 from milk to Minas cheese. In the present study, the distribution and stability of AFM1 was determined in Minas Frescal manufactured with or without starter cultures containing Lactococcus lactis spp. lactis and L. lactis spp. cremoris. 2. Material and methods 2.1. Preparation of milk containing aflatoxin M1 Raw milk used for cheese manufacture was collected at the Universidade de São Paulo, Pirassununga Campus, Brazil. Duplicate samples of milk were previously evaluated for AFM1 to confirm levels below the detection limit of the method (0.01 ng g1). For each repetition, 40 l of milk were divided into four portions of 10 l treatments containing the following: 0.250 ng AFM1 ml1, 0.500 ng AFM1 ml1, 0.250 ng AFM1 ml1 þ starter culture, and 0.500 ng AFM1 ml1 þ starter culture. Milk samples were spiked with AFM1 (SigmaeAldrich, USA) at 5.0 mg AFM1 ml1 that had been previously prepared in acetonitrile and calibrated using a spectrophotometer (l ¼ 350 nm), according to AOAC (2000) (Official Method 971.22). After calibration, 2.5 mg and 5.0 mg AFM1 were transferred to four Erlenmeyer flasks. The solvent was evaporated at room temperature under nitrogen flow. Five hundred ml of milk at 30 C was added to each Erlenmeyer and on an orbital shaker for 30 min. The total content of each flask was added to milk in vats for cheese manufacture, in order to obtain the required levels of AFM1 for treatments (0.250 and 0.500 ng AFM1 ml1). The milk Batches were continuously stirred for 30 min to ensure complete AFM1 dispersion. Milk samples (100 ml) were collected and analyzed for AFM1 to verify the levels in each treatment. 2.2. Manufacture of Minas Frescal cheese Milk was used immediately after receiving and sampling for cheese manufacture in the four treatments. The Minas Frescal cheeses were manufactured following the procedures as described by Oliveira (1986). All procedures for the manufacture of cheeses were performed in double-jacketed stainless steel vats (Mec Milk, Brazil). The raw milk (10 l per treatment) was batch pasteurized (65 C for 30 min) and then cooled to 34 C. Pasteurized milk was added with 12.5% calcium chloride solution, 12 ml of liquid rennet (chymosin and bovine pepsin, Ha-La, Christian-HansenÒ, Brazil) and salted (2%). Two treatments were inoculated with starter culture (0.5%, v/v e L. lactis spp. lactis and L. lactis spp. cremoris 1:1 e Christian HansenÒ, Brazil). Milk was incubated at 34 C for approximately 35 min and the coagulum, cut into small cubes of approximately 1.0 cm square and gently stirred for 5 min followed by manual stirring for 25 min. At the end of stirring period, whey was drained and collected for analysis. All fines in the whey were added back into the vat. The curd was then removed from the vat and transferred to molds (250 g). Cheeses were turned three times at intervals of 15 min and stored at 4 C for 24 h. Cheeses were removed from the molds, vacuum packed in polyethylene bags and stored at 4 C for 30 days. The experiment was replicated four times, representing sixteen batches of cheeses with four within treatments. 2.3. Sampling and analyses Curd samples (100 g) were collected immediately before cutting and whey samples (100 ml) were collected after mixing the whey
105
from the vat and from the cheese draining. Cheese samples (100 g) were stored at 4 C for 30 days. AFM1 analyses were performed in whey, curd and in cheeses on days 2, 9, 16, 23 and 30 after manufacture. Physical-chemical analyses of Minas Frescal cheese were performed on day 2 after manufacture and included pH, total protein and total solids evaluation (AOAC, 2000). Actual cheese yield was calculated using Y (%) ¼ (Wf/Wi) 100, where Y is the yield, Wf is the final weight of the cheese and Wi is the initial total weight (milk and ingredients). The calculated actual yield value was used for determination of the moisture-adjusted cheese yield, according to Fox, Guinee, Cogan, and McSweeney (2000). A value of 57% was used as reference cheese moisture content in Minas Frescal cheese (Oliveira, 1986). 2.4. Determination of aflatoxin M1 in milk and whey Milk and whey analysis for AFM1 were performed by the method recommended by Tuinstra, Roos, and Van Trijp (1993) with some modifications recommended by the manufacturer of the immunoaffinity columns (NeocolumnÒ e Neogen, 2005) and determined by high performance liquid chromatography (HPLC), as briefly described below. 2.4.1. Extraction and clean up 5 g of NaCl was added to each sample (40 ml; run in duplicate) was added by and then centrifuged at 3500 rpm for 10 min. The fat layer was removed and the residue filtered (1.5 mm microfiber filter, 90 mm, Millipore) and passed through an immunoaffinity column which was connected to a glass syringe and a vacuum system (flow of 2e3 ml min1). The column was washed with 20 ml of ultrapure water (Milli Q, Millipore)-methanol (75 þ 25) and the AFM1 was eluted with 1 ml of acetic acid-methanol (2 þ 98) followed by 0.4 ml ultrapure water, and the eluates were collected into an in an ambervial. 2.4.2. Chromatographic quantification of aflatoxin M1 AFM1 was quantitated by high performance liquid chromatography using a Shimadzu 10VP liquid chromatograph with a 10 AXL fluorescence detector (excitation at 366 nm and emission at 428 nm) with a Shim-Pack CLC-ODS Sil column (4.6 250 mm, 5 mm) and a Shim-Pack precolumn (4 10 mm CLC G-ODS). The isocratic mobile phase consisted of water-methanol-acetonitrile (60 þ 20þ20, v/v/v) with a flow rate of 1.0 ml min1. Under these conditions, the retention time for AFM1 was approximately 6 min (Fig. 1). A calibration curve for AFM1 was prepared using standard solutions of AFM1 (SigmaÒ), diluted in acetonitrile at concentrations of 1.25 ng ml1, 2.5 ng ml1, 5.0 ng ml1, 10.0 ng ml1 and 20.0 ng ml1 as previously described by Scott (1990),. The coefficient of determination (R2) for the AFM1 curves ranged from 0.9816 to 0.9979 attesting to the repeatability (Horwitz, 1982) of calibration curves for AFM1. 20 ml of sample extracts were injected for AFM1 determination. 2.5. Determination of aflatoxin M1 in curd and cheese Curd and cheese analyses for AFM1 determination were performed as recommended by Mayes and McDonald (1995), with some adaptations recommended by the manufacturer of the immunoaffinity columns (NeocolumnÒ e Neogen, 2005). 2.5.1. Extraction and clean up The analytical sample (10 g of each duplicate from the same sample, previously ground) was added to 5 g of celite, 150 ml chloroform and 1 ml NaCl solution the mixture was blended for
106
A.M. Fernandes et al. / Food Control 24 (2012) 104e108
with 20 ml of ultrapure water (Milli Q, Millipore)-methanol solution (75 þ 25, v/v) and the AFM1 was eluted with 1 ml methanol, into an ambervial. 2.5.2. HPLC quantification of aflatoxin M1 Identification and quantification of AFM1 in curd and cheese were achieved at the same conditions described for milk and whey. 2.6. Statistical analysis Results obtained were submitted to variance analysis in order to verify differences among treatments concerning AFM1 transfer from milk to cheeses. For evaluation of AFM1 stability in cheeses during storage SASÒ proc mixed (SAS, 2004) was used for four repetitions with five measurements, at 2, 9, 16, 23 and 30 days of storage. 3. Results and discussion The analytical methods used in the experiment were previously evaluated by the same analyst using triplicate samples of milk and whey spiked with 0.050 and 0.500 ng AFM1 ml1, and curd and cheese spiked with AFM1 at 0.200 and 0.500 ng g1. The limit of determination (LOD) and limit of quantification (LOQ) for AFM1 were determined in all types of samples considering a signal to noise ratio of 3:1 and 6:1, respectively. The LOD and LOQ values for AFM1 in milk and whey were 0.005 ng ml1 and 0.010 ng ml1, respectively. For AFM1 in curd and cheese samples, the LOD and LOQ were 0.010 ng g1 and 0.030 ng g1, respectively. The mean recoveries of AFM1 in spiked milk or whey samples were in the range 75e94%, with relative standard deviation (RSD) from 4.46 to 10.49%. For curd or cheese samples, recovery values varied from 53 to 84%, with RSD of 3.61e12.80%. Similar values for recoveries in milk (90e100%) and Camembert and St Albray cheeses (72e90%) have been described by Tuinstra et al. (1993) and Dragacci, Gleizes, Fremy, and Candlish (1995), respectively. Reproducibility (RSDR) was not evaluated in the present study, although the method used for cheeses presented RSDR values ranging from 11 to 23% in an inter-laboratorial as reported by Tuinstra et al. (1993). The yield and physical-chemical results obtained for Minas Frescal cheeses did not present significant differences (P > 0.05) among means. Moisture-adjusted cheese yield ranged from 11.97 to 14.69%, total protein varied from 12.45 to 13.97%, total solids from 31.27 to 32.73%, and pH mean was 6.4 (data not shown). The concentrations of AFM1 and its percentages of carrying-over from contaminated milks into whey, curd and Minas Frescal cheeses are presented in Table 1. AFM1 levels ranged from 0.099 to
Fig. 1. (A) Chromatogram showing the retention time of aflatoxin M1 (AFM1) (nearly 6 min) in a standard with 10 ng AFM1 ml1; (B) Chromatogram of an artificially contaminated (500 ng AFM1 g1) Minas Frescal cheese sample at 30 days of storage.
2e3 min, filtered and evaporated. The residue was dissolved in 2 ml of methanol and the volume adjusted to 100 ml using phosphatesalt buffer, prepared according to Deveci (2007). The solution was placed into a separation funnel with 75 ml of n-hexane and mixed. The aqueous phase was separated and passed through an immunoaffinity column which was connected to a glass syringe and a vacuum system (flow of 2e3 ml min1). The column was washed
Table 1 Aflatoxin M1 concentrations in milk, whey, curd and cheese from different treatments and recoveries.a Treatment
Milk
Whey 1
AFM1 (ng ml
)
Cheeseb
Curd 1
Total AFM1 (ng)
AFM1 (ng ml
0.250 ng ml1 Without SC 0.230 0.037b With SC 0.234 0.087b
2.630 930b 2.624 519b
0.500 ng ml1 Without SC 0.448 0.016a With SC 0.502 0.106a
5.017 1309a 5.586 1323a
)
c
Recovery (%)
AFM1 (ng g1)
Concentration factord
0.199 0.056a 0.210 0.046a
42.26 13.28a 30.64 8.65a
0.540 0.132b 0.432 0.120b
2.45 0.92 1.84 0.49
0.416 0.142a 0.391 0.162a
38.51 10.99a 34.91 16.57a
0.778 0.159a 0.819 0.285a
1.79 0.52 1.74 0.91
Recovery (%)
AFM1 (ng g
0.099 0.011b 0.102 0.021b
39.41 10.36a 39.71 9.58a
0.215 0.040a 0.196 0.051a
41.91 11.24a 35.57 15.10a
1
)
In the same column, means followed by different letters present significant difference (P < 0.05). SC: Starter culture (Lactococcus lactis spp. lactis and Lactococcus lactis spp. cremoris 1:1 e Christian HansenÒ, Brazil). a Results are expressed as mean standard deviation. N ¼ 4. b Day 2 after manufacture. c Percentage of AFM1 found in relation to the total AFM1 in the original milk. d Level of AFM1 in cheese divided by the level in the original milk.
A.M. Fernandes et al. / Food Control 24 (2012) 104e108 1.600
AFM1 (ng ml-1)
1.400 1.200 1.000 0.800 0.600 0.400 0.200 0.000 2
9
16
23
30
Days of storage 0.250 ng mL-1 wit hout SC
0.250 ng mL-1 wit h SC
0.500 ng mL-1 wit hout SC
0.500 ng mL-1 wit h SC
Fig. 2. Levels (mean standard deviation, N ¼ 4) of AFM1 in Minas Frescal cheese during 30 days of storage. There was no effect of time of storage (P > 0.05) and there was no interaction between AFM1 and time of storage (P > 0.05). SC: Starter culture (Lactococcus lactis spp. lactis and L. lactis spp. cremoris 1:1 e Christian HansenÒ, Brazil).
0.215 ng ml1 in whey, and from 0.199 to 0.416 ng g1 in curd. Cheese on day 2 after manufacture had levels of AFM1 from 0.432 to 0.819 ng g1. There were no significant differences (P > 0.05) among treatment means concerning AFM1 transfer from milk. AFM1 transferred into whey had 39.41% when the milk contained 0.250 ng AFM1 ml1 and 41.91% when added by 0.500 ng AFM1 ml1, without use of starter culture, and 39.71% and 35.57% when added by 0.250 and 0.500 ng AFM1 ml1, respectively, with starter culture. Minas Frescal cheeses had 42.26% and 38.51% of AFM1 transferred from milk in treatments 0.250 and 0.500 ng AFM1 ml1, respectively, without culture. Treatments 0.250 and 0.500 ng AFM1 ml1 with culture had AFM1 transfer from milk to cheese at 30.64% and 34.91%, respectively. White cheeses from North Africa produced by using milk artificially contaminated with AFM1 in concentrations of 0.01e3.0 ng ml1 had AFM1 concentrations ranging from 0.11 to 0.52 ng g1 (Elgerbi et al., 2004). White cheese from Iran had AFM1 levels of 0.22, 0.88 and 0.75 ng ml1 in whey, curd and cheese, respectively, when milk was artificially contaminated with 0.50 ng AFM1 ml1 (Kamkar, Karim, Shojaee Aliabadi, & Khaksar, 2008). Fremy, Roiland, and Gaymard (1990) evaluated Camembert cheese produced with milk artificially contaminated with AFM1 at levels as high as 0.3e7.5 ng ml1 and observed transfers of 35.6 and 57.7% of AFM1, respectively, from milk to cheese, and 42.3 and 64.4%, respectively, to whey. Fresh cheese produced in Argentina using milk artificially contaminated with AFM1 at levels of 1.7e2.0 ng ml1 had 60% of the AFM1 in whey and 40% in cheese (López et al., 2001). The use of starter culture did not influence AFM1 transfer from milk to cheese (P > 0.05) (Table 1). Oruc, Cibik, Yilmaz, and Gunes (2007) also found cultures did not influence the amount of AFM1 in Kashar cheese attributing this fact to the high pH values of cheese, around 5.0e5.2. Total recovery (whey and cheese) of AFM1 added to milk were 81.67% 19.00, 80.42% 21.99, 70.35% 13.08 and 70.48% 31.47 for treatments without culture 0.250 and 0.500 and with culture 0.250 and 0.500 ng AFM1 ml1, respectively. After 2 days of storage, AFM1 in cheeses was at levels of 1.74- to 2.45-fold higher than those presented in the initial artificially contaminated milk (Table 1). On the 30th day of storage, AFM1 was 2.14- to 2.60-fold higher than levels in milk. There was concentration of the toxin in cheese, independently (P > 0.3443) of AFM1 level in milk (0.250 or 0.500 ng ml1). The concentration of AFM1 in Minas Frescal cheese relative to milk was lower than those reported by researchers for other types of cheese. Parmesan and Mozzarella found AFM1 levels of 5.8- and 7.1-fold higher than in milk (Brackett
107
& Marth, 1982), while the level in Cottage cheese was 8.1-fold higher (Applebaum et al., 1982). Different types of cheese produced with milk artificially contaminated with AFM1 have been reported to have concentrations 1.8- to 4.4-fold higher than in milk (Govaris, Roussi, Koidis, & Botsoglou, 2001; Oruc et al., 2006; Kamkar et al., 2008). Over 30 days of storage (Fig. 2), AFM1 levels were constant (P ¼ 0.2777), representing no degradation and no losses by whey of AFM1 during 30 days of storage. These results are in agreement with those reported by Oruc et al. (2006), in which AFM1 remained stable during ripening in white Turkish cheese. However, some types of cheeses, such as Camembert and Telemes, had a decrease in AFM1 concentrations during storage (Fremy et al., 1990; Govaris et al., 2001). In our study, the use of starter cultures did not influence (P ¼ 0.2083) AFM1 stability for 30 days. 4. Conclusion The carry-over of AFM1 from milk to Minas Frescal was 30e40% of the amount in the original milk. The use of milk containing 0.250e0.500 ng AFM1 ml1 may concentrate the level of AFM1 in Minas Frescal cheeses up to 2.5-fold, but at levels below the tolerance limit for AFM1 in cheeses (2.5 ng g1) as recently established by Brazilian regulations (Agência Nacional de Vigilância Sanitária, 2011). Also, the addition of starter cultures in Minas Frescal cheese as used in this trial did not influence the concentration or the stability of the AFM1 in cheeses over 30 days storage. Acknowledgments The authors thank the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Brazil e grant no 2006/61580-0, for financial support. References Agência Nacional de Vigilância Sanitária. Resolução RDC n 7, de 18 de fevereiro de 2011. Diário Oficial da União e Seção 1, n 37, 22 de fevereiro de 2011. pp. 72e73, 2011. AOAC. (2000). Association of official analytical chemists. Official methods of analysis (17th ed.). Arlington, VA: AOAC. Applebaum, R. S., Brackett, R. E., Wiseman, D. W., & Marth, E. H. (1982). Aflatoxin: toxicity to dairy cattle and occurrence in milk and milk products e a review. Journal of Food Protection, 45, 752e777. Brackett, R. E., & Marth, E. H. (1982). Association of aflatoxin M1 with casein. Zeitschrift fur Lebensmittel-Untersuchung und Forschung, 174, 439e441. Deveci, O. (2007). Changes in the concentration of aflatoxin M1 during manufacture and storage of White Pickled cheese. Food Control, 18, 1103e1107. Dragacci, S., Gleizes, E., Fremy, J. M., & Candlish, A. A. G. (1995). Use of immunoaffinity chromatography as a purification step for the determination of aflatoxin M1 in cheeses. Food Additives and Contaminants, 12, 59e65. Elgerbi, A. M., Aidoo, K. E., Candlish, A. A. G., & Tester, R. F. (2004). Occurrence of aflatoxin M1 in randomly selected North African milk and cheese samples. Food Additives and Contaminants, 21, 592e597. Food and Agriculture Organization. (2004). Worldwide regulations for mycotoxins in food and feed in 2003. Rome: Food and Agriculture Organization. (FAO Food and Nutrition Paper, 81). Fox, P. F., Guinee, T. P., Cogan, T. M., & McSweeney, P. L. H. (2000). Fundamentals of cheese science. Gaithersburg: Aspen Publishers. Fremy, J. M., Roiland, J. C., & Gaymard, A. (1990). Behavior of 14C aflatoxin M1 during Camembert cheese making. Journal of Environmental Pathology, Toxicology and Oncology, 10, 95e98. Govaris, A., Roussi, V., Koidis, P. A., & Botsoglou, N. A. (2001). Distribution and stability of aflatoxin M1 during processing, ripening and storage of Telemes cheese. Food Additives and Contaminants, 18, 437e443. Horwitz, W. (1982). Evaluation of analytical methods used for regulation of foods and drugs. Analytical Chemistry, 54, 67Ae76A. International Agency for Research on Cancer. (1993). Some Naturally occurring substances: food Items and constituents, heterocyclic aromatic amines and mycotoxins. In. Monographs on the evaluation of carcinogenic risks to humans, Vol. 56 (pp. 489e521). Lyon: IARC. Jouany, J. P., & Diaz, D. E. (2005). Effects of mycotoxins in ruminants. In D. E. Diaz (Ed.), The mycotoxin blue book (pp. 295e321). Nottingham: Nottingham University Press. Kamkar, A., Karim, G., Shojaee Aliabadi, F., & Khaksar, R. (2008). Fate of aflatoxin M1 in Iranian white cheese processing. Food and Chemical Toxicology, 46, 2236e2238.
108
A.M. Fernandes et al. / Food Control 24 (2012) 104e108
López, C., Ramos, L., Ramadan, S., Bulacio, L., & Perez, J. (2001). Distribution of aflatoxin M1 in cheese obtained from milk artificially contaminated. International Journal of Food Microbiology, 64, 211e215. Mayes, L., & McDonald, S. (1995). Aflatoxin M1 in retail milk and milk products. MAFF. CSL Food Science Laboratory. Report No. FD 94/98a. Moss, M. O. (1998). Recent studies of mycotoxins. Journal of Applied Microbioogy e Symposium Supplement, 84, 62Se76S. Neogen, Europe Ltd. (2005). Neocolumn instruction manual. Scotland, U.K. Oliveira, J. S. (1986). Queijo: fundamentos tecnológicos (2a ed). Campinas: UNICAMP. Oliveira, C. A. F., Franco, R. C., Rosim, R. E., & Fernandes, A. M. (2011). Survey of aflatoxin M1 in Minas cheese from the North-east region of São Paulo, Brazil. Food Additives and Contaminants: Part B, 4, 57e60. Oruc, H. H., Cibik, R., Yilmaz, E., & Kalkanli, O. (2006). Distribution and stability of Aflatoxin M1 during processing and ripening of traditional white pickled cheese. Food Additives and Contaminants, 23, 190e195. Oruc, H. H., Cibik, R., Yilmaz, E., & Gunes, E. (2007). Fate of aflatoxin M1 in Kashar cheese. Journal of Food Safety, 27, 82e90.
Park, D. L., & Pohland, A. E. (1986). A rationale for the control of aflatoxin in animal feeds. In P. S. Steyn, & R. Vleggaar (Eds.), Mycotoxins and phycotoxins (pp. 473e482). Amsterdam: Elsevier Applied Science. Pierides, M., El-Nizami, H., Peltonen, K., Salminen, S., & Ahokas, J. (2000). Ability of dairy strains of lactic acid bacteria to bind aflatoxin M1 in a food model. Journal of Food Protection, 63, 645e650. Prado, G., Oliveira, M. S., Pereira, M. L., Abrantes, F. M., Santos, L. G., & Veloso, T. (2000). Aflatoxin M1 in samples of “Minas” cheese commercialized in the city of Belo Horizonte e Minas Gerais/Brazil. Ciência e Tecnologia de Alimentos, 20, 398e400. SAS Institute. (2004). SASÒ user’s guide statistics. Cary, NC: SAS Institute Inc. Scott, P. M. (1990). Natural poisons. In K. Helrich (Ed.), Official methods of analysis of the Association of Official Analytical Chemists (15th ed). (pp. 1184e1213). Arlington: AOAC. Tuinstra, L. G., Roos, A. H., & Van Trijp, J. M. (1993). Liquid chromatographic determination of aflatoxin M1 in milk powder using immunoaffinity columns for clean-up: interlaboratory study. Food Biology Contamination, 76, 1248e1254. Wood, G. E. (1991). Aflatoxin M1. In R. P. Sharma, & D. K. Salunkhe (Eds.), Mycotoxins and phytoalexins (pp. 145e164). Boca Raton: CRC Press.
Contents lists available at SciVerse ScienceDirect
Food Control journal homepage: www.elsevier.com/locate/foodcont
Distribution and stability of aflatoxin M1 during processing and storage of Minas Frescal cheese A.M. Fernandesa, B. Corrêab, R.E. Rosimc, E. Kobashigawac, C.A.F. Oliveirac, * a
Department of Basic Sciences, School of Animal Science and Food Engineering, University of São Paulo, Av. Duque de Caxias Norte, 225, CEP 13635-900, Pirassununga, SP, Brazil Department of Microbiology, Institute of Biomedical Sciences, University of São Paulo, Av. Prof. Lineu Prestes, 1374, CEP 05508-900, São Paulo, SP, Brazil c Department of Food Engineering, School of Animal Science and Food Engineering, University of São Paulo, Av. Duque de Caxias Norte, 225, CEP 13635-900, Pirassununga, SP, Brazil b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 May 2011 Received in revised form 12 August 2011 Accepted 6 September 2011
Aflatoxin M1 (AFM1) is a hepatocarcinogen found in milk of animals that have consumed feeds with aflatoxin B1. The carry-over of AFM1 from milk to Minas Frescal cheese produced with or without starter cultures was determined. 40 L of milk were divided into 10 L each and assigned to the following treatments for cheese manufacture: 0.250 ng AFM1 mL1, 0.500 ng AFM1 mL1, 0.250 ng AFM1 mL1 þ starter, 0.500 ng AFM1 mL1 þ starter. Quantification of AFM1 was achieved by high performance liquid chromatography. The carry-over of AFM1 from milk to cheese ranged from 30.64% to 42.26%. There was no effect of storage time on AFM1. Milk with AFM1 in levels studied may concentrate the toxin in Minas Frescal cheese, but at concentrations below the Brazilian tolerance limit. The addition of starter cultures did not influence concentration or stability of the AFM1 in cheese over 30 days storage. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Mycotoxins AFM1 Cheese Residues
1. Introduction Aflatoxin M1 (AFM1) is a hepatocarcinogenic metabolite found in milk of animals that have consumed feeds contaminated with aflatoxin B1, AFB1 (Park & Pohland, 1986), the major mycotoxin produced by fungi of the genus Aspergillus (Moss, 1998). The carryover of AFM1 into milk has been demonstrated in several studies (Wood, 1991) and the rate of transfer of AFB1 from feed to milk in dairy cows is between 1 and 3% with a mean transfer of 1.7% (Jouany & Diaz, 2005). The occurrence of AFM1 in milk and dairy products is a public health concern since the International Agency for Research on Cancer (1993) has classified AFM1 as a probable human carcinogen (Group 2). There are several reports of AFM1 contamination of cheeses produced from milk artificially treated with AFB1; however, these reports refer to major cheeses from each country and, according to López, Ramos, Ramadan, Bulacio, and Perez (2001), the AFM1 levels in cheese seem to be variable, depending on the type of the product. Studies involving different types of cheeses demonstrated the presence of higher levels of AFM1 in cheese than found in the artificially contaminated milk (Elgerbi, Aidoo, Candlish, & Tester, 2004; Oruc, Cibik, Yilmaz, & Kalkanli, 2006). The distribution of
* Corresponding author. Tel.: þ55 19 3565 4173; fax: þ55 19 3565 4284. E-mail address: [email protected] (C.A.F. Oliveira). 0956-7135/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.foodcont.2011.09.010
AFM1 during cheese manufacture also varies with the type of final product. Fresh cheese produced in Argentina from artificially contaminated milk with AFM1 at levels of 1.7e2.0 ng mL1 had 60% of AFM1 in the whey and 40% in cheese (López et al., 2001). In Brazil, there is little information on the occurrence of AFM1 in cheeses manufactured in the country, except for two reports, one of them by Prado et al. (2000), who observed that 74.7% of Minas Frescal samples from the state of Minas Gerais showed AFM1 at levels ranging from 0.02 to 6.92 ng g1. In a recent study, Oliveira, Franco, Rosim, and Fernandes (2011) found a high incidence of AFM1 in Minas Frescal cheeses produced in the state of São Paulo, but levels were usually below the tolerance limit (2.5 ng g1) established by Brazilian regulations (Agência Nacional de Vigilância Sanitária, 2011). Other countries have established different tolerance limits for AFM1 in cheeses: 0.20 ng g1 in the Netherlands, and 0.25 ng g1 in Austria, Switzerland and Turkey (Food and Agriculture Organization, 2004). Several studies have suggested that starter culture bacteria used in cheese manufacture might inactivate AFM1 by forming a linkage between AFM and lactic bacteria (Pierides, El-Nizami, Peltonen, Salminen, & Ahokas, 2000). In addition, AFM1 may form linkages with milk proteins such as casein depending on the influence of acidity (Brackett & Marth, 1982). Minas Frescal cheese is typical Brazilian cheese classified as a semi-fat, non-ripened and high humidity cheese which is obtained by enzymatic coagulation of milk with rennet and with or without starter cultures. Although in
A.M. Fernandes et al. / Food Control 24 (2012) 104e108
Brazil previous studies indicated a high incidence of AFM1 in fluid milk, there are no previous reports on the carry-over of AFM1 from milk to Minas cheese. In the present study, the distribution and stability of AFM1 was determined in Minas Frescal manufactured with or without starter cultures containing Lactococcus lactis spp. lactis and L. lactis spp. cremoris. 2. Material and methods 2.1. Preparation of milk containing aflatoxin M1 Raw milk used for cheese manufacture was collected at the Universidade de São Paulo, Pirassununga Campus, Brazil. Duplicate samples of milk were previously evaluated for AFM1 to confirm levels below the detection limit of the method (0.01 ng g1). For each repetition, 40 l of milk were divided into four portions of 10 l treatments containing the following: 0.250 ng AFM1 ml1, 0.500 ng AFM1 ml1, 0.250 ng AFM1 ml1 þ starter culture, and 0.500 ng AFM1 ml1 þ starter culture. Milk samples were spiked with AFM1 (SigmaeAldrich, USA) at 5.0 mg AFM1 ml1 that had been previously prepared in acetonitrile and calibrated using a spectrophotometer (l ¼ 350 nm), according to AOAC (2000) (Official Method 971.22). After calibration, 2.5 mg and 5.0 mg AFM1 were transferred to four Erlenmeyer flasks. The solvent was evaporated at room temperature under nitrogen flow. Five hundred ml of milk at 30 C was added to each Erlenmeyer and on an orbital shaker for 30 min. The total content of each flask was added to milk in vats for cheese manufacture, in order to obtain the required levels of AFM1 for treatments (0.250 and 0.500 ng AFM1 ml1). The milk Batches were continuously stirred for 30 min to ensure complete AFM1 dispersion. Milk samples (100 ml) were collected and analyzed for AFM1 to verify the levels in each treatment. 2.2. Manufacture of Minas Frescal cheese Milk was used immediately after receiving and sampling for cheese manufacture in the four treatments. The Minas Frescal cheeses were manufactured following the procedures as described by Oliveira (1986). All procedures for the manufacture of cheeses were performed in double-jacketed stainless steel vats (Mec Milk, Brazil). The raw milk (10 l per treatment) was batch pasteurized (65 C for 30 min) and then cooled to 34 C. Pasteurized milk was added with 12.5% calcium chloride solution, 12 ml of liquid rennet (chymosin and bovine pepsin, Ha-La, Christian-HansenÒ, Brazil) and salted (2%). Two treatments were inoculated with starter culture (0.5%, v/v e L. lactis spp. lactis and L. lactis spp. cremoris 1:1 e Christian HansenÒ, Brazil). Milk was incubated at 34 C for approximately 35 min and the coagulum, cut into small cubes of approximately 1.0 cm square and gently stirred for 5 min followed by manual stirring for 25 min. At the end of stirring period, whey was drained and collected for analysis. All fines in the whey were added back into the vat. The curd was then removed from the vat and transferred to molds (250 g). Cheeses were turned three times at intervals of 15 min and stored at 4 C for 24 h. Cheeses were removed from the molds, vacuum packed in polyethylene bags and stored at 4 C for 30 days. The experiment was replicated four times, representing sixteen batches of cheeses with four within treatments. 2.3. Sampling and analyses Curd samples (100 g) were collected immediately before cutting and whey samples (100 ml) were collected after mixing the whey
105
from the vat and from the cheese draining. Cheese samples (100 g) were stored at 4 C for 30 days. AFM1 analyses were performed in whey, curd and in cheeses on days 2, 9, 16, 23 and 30 after manufacture. Physical-chemical analyses of Minas Frescal cheese were performed on day 2 after manufacture and included pH, total protein and total solids evaluation (AOAC, 2000). Actual cheese yield was calculated using Y (%) ¼ (Wf/Wi) 100, where Y is the yield, Wf is the final weight of the cheese and Wi is the initial total weight (milk and ingredients). The calculated actual yield value was used for determination of the moisture-adjusted cheese yield, according to Fox, Guinee, Cogan, and McSweeney (2000). A value of 57% was used as reference cheese moisture content in Minas Frescal cheese (Oliveira, 1986). 2.4. Determination of aflatoxin M1 in milk and whey Milk and whey analysis for AFM1 were performed by the method recommended by Tuinstra, Roos, and Van Trijp (1993) with some modifications recommended by the manufacturer of the immunoaffinity columns (NeocolumnÒ e Neogen, 2005) and determined by high performance liquid chromatography (HPLC), as briefly described below. 2.4.1. Extraction and clean up 5 g of NaCl was added to each sample (40 ml; run in duplicate) was added by and then centrifuged at 3500 rpm for 10 min. The fat layer was removed and the residue filtered (1.5 mm microfiber filter, 90 mm, Millipore) and passed through an immunoaffinity column which was connected to a glass syringe and a vacuum system (flow of 2e3 ml min1). The column was washed with 20 ml of ultrapure water (Milli Q, Millipore)-methanol (75 þ 25) and the AFM1 was eluted with 1 ml of acetic acid-methanol (2 þ 98) followed by 0.4 ml ultrapure water, and the eluates were collected into an in an ambervial. 2.4.2. Chromatographic quantification of aflatoxin M1 AFM1 was quantitated by high performance liquid chromatography using a Shimadzu 10VP liquid chromatograph with a 10 AXL fluorescence detector (excitation at 366 nm and emission at 428 nm) with a Shim-Pack CLC-ODS Sil column (4.6 250 mm, 5 mm) and a Shim-Pack precolumn (4 10 mm CLC G-ODS). The isocratic mobile phase consisted of water-methanol-acetonitrile (60 þ 20þ20, v/v/v) with a flow rate of 1.0 ml min1. Under these conditions, the retention time for AFM1 was approximately 6 min (Fig. 1). A calibration curve for AFM1 was prepared using standard solutions of AFM1 (SigmaÒ), diluted in acetonitrile at concentrations of 1.25 ng ml1, 2.5 ng ml1, 5.0 ng ml1, 10.0 ng ml1 and 20.0 ng ml1 as previously described by Scott (1990),. The coefficient of determination (R2) for the AFM1 curves ranged from 0.9816 to 0.9979 attesting to the repeatability (Horwitz, 1982) of calibration curves for AFM1. 20 ml of sample extracts were injected for AFM1 determination. 2.5. Determination of aflatoxin M1 in curd and cheese Curd and cheese analyses for AFM1 determination were performed as recommended by Mayes and McDonald (1995), with some adaptations recommended by the manufacturer of the immunoaffinity columns (NeocolumnÒ e Neogen, 2005). 2.5.1. Extraction and clean up The analytical sample (10 g of each duplicate from the same sample, previously ground) was added to 5 g of celite, 150 ml chloroform and 1 ml NaCl solution the mixture was blended for
106
A.M. Fernandes et al. / Food Control 24 (2012) 104e108
with 20 ml of ultrapure water (Milli Q, Millipore)-methanol solution (75 þ 25, v/v) and the AFM1 was eluted with 1 ml methanol, into an ambervial. 2.5.2. HPLC quantification of aflatoxin M1 Identification and quantification of AFM1 in curd and cheese were achieved at the same conditions described for milk and whey. 2.6. Statistical analysis Results obtained were submitted to variance analysis in order to verify differences among treatments concerning AFM1 transfer from milk to cheeses. For evaluation of AFM1 stability in cheeses during storage SASÒ proc mixed (SAS, 2004) was used for four repetitions with five measurements, at 2, 9, 16, 23 and 30 days of storage. 3. Results and discussion The analytical methods used in the experiment were previously evaluated by the same analyst using triplicate samples of milk and whey spiked with 0.050 and 0.500 ng AFM1 ml1, and curd and cheese spiked with AFM1 at 0.200 and 0.500 ng g1. The limit of determination (LOD) and limit of quantification (LOQ) for AFM1 were determined in all types of samples considering a signal to noise ratio of 3:1 and 6:1, respectively. The LOD and LOQ values for AFM1 in milk and whey were 0.005 ng ml1 and 0.010 ng ml1, respectively. For AFM1 in curd and cheese samples, the LOD and LOQ were 0.010 ng g1 and 0.030 ng g1, respectively. The mean recoveries of AFM1 in spiked milk or whey samples were in the range 75e94%, with relative standard deviation (RSD) from 4.46 to 10.49%. For curd or cheese samples, recovery values varied from 53 to 84%, with RSD of 3.61e12.80%. Similar values for recoveries in milk (90e100%) and Camembert and St Albray cheeses (72e90%) have been described by Tuinstra et al. (1993) and Dragacci, Gleizes, Fremy, and Candlish (1995), respectively. Reproducibility (RSDR) was not evaluated in the present study, although the method used for cheeses presented RSDR values ranging from 11 to 23% in an inter-laboratorial as reported by Tuinstra et al. (1993). The yield and physical-chemical results obtained for Minas Frescal cheeses did not present significant differences (P > 0.05) among means. Moisture-adjusted cheese yield ranged from 11.97 to 14.69%, total protein varied from 12.45 to 13.97%, total solids from 31.27 to 32.73%, and pH mean was 6.4 (data not shown). The concentrations of AFM1 and its percentages of carrying-over from contaminated milks into whey, curd and Minas Frescal cheeses are presented in Table 1. AFM1 levels ranged from 0.099 to
Fig. 1. (A) Chromatogram showing the retention time of aflatoxin M1 (AFM1) (nearly 6 min) in a standard with 10 ng AFM1 ml1; (B) Chromatogram of an artificially contaminated (500 ng AFM1 g1) Minas Frescal cheese sample at 30 days of storage.
2e3 min, filtered and evaporated. The residue was dissolved in 2 ml of methanol and the volume adjusted to 100 ml using phosphatesalt buffer, prepared according to Deveci (2007). The solution was placed into a separation funnel with 75 ml of n-hexane and mixed. The aqueous phase was separated and passed through an immunoaffinity column which was connected to a glass syringe and a vacuum system (flow of 2e3 ml min1). The column was washed
Table 1 Aflatoxin M1 concentrations in milk, whey, curd and cheese from different treatments and recoveries.a Treatment
Milk
Whey 1
AFM1 (ng ml
)
Cheeseb
Curd 1
Total AFM1 (ng)
AFM1 (ng ml
0.250 ng ml1 Without SC 0.230 0.037b With SC 0.234 0.087b
2.630 930b 2.624 519b
0.500 ng ml1 Without SC 0.448 0.016a With SC 0.502 0.106a
5.017 1309a 5.586 1323a
)
c
Recovery (%)
AFM1 (ng g1)
Concentration factord
0.199 0.056a 0.210 0.046a
42.26 13.28a 30.64 8.65a
0.540 0.132b 0.432 0.120b
2.45 0.92 1.84 0.49
0.416 0.142a 0.391 0.162a
38.51 10.99a 34.91 16.57a
0.778 0.159a 0.819 0.285a
1.79 0.52 1.74 0.91
Recovery (%)
AFM1 (ng g
0.099 0.011b 0.102 0.021b
39.41 10.36a 39.71 9.58a
0.215 0.040a 0.196 0.051a
41.91 11.24a 35.57 15.10a
1
)
In the same column, means followed by different letters present significant difference (P < 0.05). SC: Starter culture (Lactococcus lactis spp. lactis and Lactococcus lactis spp. cremoris 1:1 e Christian HansenÒ, Brazil). a Results are expressed as mean standard deviation. N ¼ 4. b Day 2 after manufacture. c Percentage of AFM1 found in relation to the total AFM1 in the original milk. d Level of AFM1 in cheese divided by the level in the original milk.
A.M. Fernandes et al. / Food Control 24 (2012) 104e108 1.600
AFM1 (ng ml-1)
1.400 1.200 1.000 0.800 0.600 0.400 0.200 0.000 2
9
16
23
30
Days of storage 0.250 ng mL-1 wit hout SC
0.250 ng mL-1 wit h SC
0.500 ng mL-1 wit hout SC
0.500 ng mL-1 wit h SC
Fig. 2. Levels (mean standard deviation, N ¼ 4) of AFM1 in Minas Frescal cheese during 30 days of storage. There was no effect of time of storage (P > 0.05) and there was no interaction between AFM1 and time of storage (P > 0.05). SC: Starter culture (Lactococcus lactis spp. lactis and L. lactis spp. cremoris 1:1 e Christian HansenÒ, Brazil).
0.215 ng ml1 in whey, and from 0.199 to 0.416 ng g1 in curd. Cheese on day 2 after manufacture had levels of AFM1 from 0.432 to 0.819 ng g1. There were no significant differences (P > 0.05) among treatment means concerning AFM1 transfer from milk. AFM1 transferred into whey had 39.41% when the milk contained 0.250 ng AFM1 ml1 and 41.91% when added by 0.500 ng AFM1 ml1, without use of starter culture, and 39.71% and 35.57% when added by 0.250 and 0.500 ng AFM1 ml1, respectively, with starter culture. Minas Frescal cheeses had 42.26% and 38.51% of AFM1 transferred from milk in treatments 0.250 and 0.500 ng AFM1 ml1, respectively, without culture. Treatments 0.250 and 0.500 ng AFM1 ml1 with culture had AFM1 transfer from milk to cheese at 30.64% and 34.91%, respectively. White cheeses from North Africa produced by using milk artificially contaminated with AFM1 in concentrations of 0.01e3.0 ng ml1 had AFM1 concentrations ranging from 0.11 to 0.52 ng g1 (Elgerbi et al., 2004). White cheese from Iran had AFM1 levels of 0.22, 0.88 and 0.75 ng ml1 in whey, curd and cheese, respectively, when milk was artificially contaminated with 0.50 ng AFM1 ml1 (Kamkar, Karim, Shojaee Aliabadi, & Khaksar, 2008). Fremy, Roiland, and Gaymard (1990) evaluated Camembert cheese produced with milk artificially contaminated with AFM1 at levels as high as 0.3e7.5 ng ml1 and observed transfers of 35.6 and 57.7% of AFM1, respectively, from milk to cheese, and 42.3 and 64.4%, respectively, to whey. Fresh cheese produced in Argentina using milk artificially contaminated with AFM1 at levels of 1.7e2.0 ng ml1 had 60% of the AFM1 in whey and 40% in cheese (López et al., 2001). The use of starter culture did not influence AFM1 transfer from milk to cheese (P > 0.05) (Table 1). Oruc, Cibik, Yilmaz, and Gunes (2007) also found cultures did not influence the amount of AFM1 in Kashar cheese attributing this fact to the high pH values of cheese, around 5.0e5.2. Total recovery (whey and cheese) of AFM1 added to milk were 81.67% 19.00, 80.42% 21.99, 70.35% 13.08 and 70.48% 31.47 for treatments without culture 0.250 and 0.500 and with culture 0.250 and 0.500 ng AFM1 ml1, respectively. After 2 days of storage, AFM1 in cheeses was at levels of 1.74- to 2.45-fold higher than those presented in the initial artificially contaminated milk (Table 1). On the 30th day of storage, AFM1 was 2.14- to 2.60-fold higher than levels in milk. There was concentration of the toxin in cheese, independently (P > 0.3443) of AFM1 level in milk (0.250 or 0.500 ng ml1). The concentration of AFM1 in Minas Frescal cheese relative to milk was lower than those reported by researchers for other types of cheese. Parmesan and Mozzarella found AFM1 levels of 5.8- and 7.1-fold higher than in milk (Brackett
107
& Marth, 1982), while the level in Cottage cheese was 8.1-fold higher (Applebaum et al., 1982). Different types of cheese produced with milk artificially contaminated with AFM1 have been reported to have concentrations 1.8- to 4.4-fold higher than in milk (Govaris, Roussi, Koidis, & Botsoglou, 2001; Oruc et al., 2006; Kamkar et al., 2008). Over 30 days of storage (Fig. 2), AFM1 levels were constant (P ¼ 0.2777), representing no degradation and no losses by whey of AFM1 during 30 days of storage. These results are in agreement with those reported by Oruc et al. (2006), in which AFM1 remained stable during ripening in white Turkish cheese. However, some types of cheeses, such as Camembert and Telemes, had a decrease in AFM1 concentrations during storage (Fremy et al., 1990; Govaris et al., 2001). In our study, the use of starter cultures did not influence (P ¼ 0.2083) AFM1 stability for 30 days. 4. Conclusion The carry-over of AFM1 from milk to Minas Frescal was 30e40% of the amount in the original milk. The use of milk containing 0.250e0.500 ng AFM1 ml1 may concentrate the level of AFM1 in Minas Frescal cheeses up to 2.5-fold, but at levels below the tolerance limit for AFM1 in cheeses (2.5 ng g1) as recently established by Brazilian regulations (Agência Nacional de Vigilância Sanitária, 2011). Also, the addition of starter cultures in Minas Frescal cheese as used in this trial did not influence the concentration or the stability of the AFM1 in cheeses over 30 days storage. Acknowledgments The authors thank the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Brazil e grant no 2006/61580-0, for financial support. References Agência Nacional de Vigilância Sanitária. Resolução RDC n 7, de 18 de fevereiro de 2011. Diário Oficial da União e Seção 1, n 37, 22 de fevereiro de 2011. pp. 72e73, 2011. AOAC. (2000). Association of official analytical chemists. Official methods of analysis (17th ed.). Arlington, VA: AOAC. Applebaum, R. S., Brackett, R. E., Wiseman, D. W., & Marth, E. H. (1982). Aflatoxin: toxicity to dairy cattle and occurrence in milk and milk products e a review. Journal of Food Protection, 45, 752e777. Brackett, R. E., & Marth, E. H. (1982). Association of aflatoxin M1 with casein. Zeitschrift fur Lebensmittel-Untersuchung und Forschung, 174, 439e441. Deveci, O. (2007). Changes in the concentration of aflatoxin M1 during manufacture and storage of White Pickled cheese. Food Control, 18, 1103e1107. Dragacci, S., Gleizes, E., Fremy, J. M., & Candlish, A. A. G. (1995). Use of immunoaffinity chromatography as a purification step for the determination of aflatoxin M1 in cheeses. Food Additives and Contaminants, 12, 59e65. Elgerbi, A. M., Aidoo, K. E., Candlish, A. A. G., & Tester, R. F. (2004). Occurrence of aflatoxin M1 in randomly selected North African milk and cheese samples. Food Additives and Contaminants, 21, 592e597. Food and Agriculture Organization. (2004). Worldwide regulations for mycotoxins in food and feed in 2003. Rome: Food and Agriculture Organization. (FAO Food and Nutrition Paper, 81). Fox, P. F., Guinee, T. P., Cogan, T. M., & McSweeney, P. L. H. (2000). Fundamentals of cheese science. Gaithersburg: Aspen Publishers. Fremy, J. M., Roiland, J. C., & Gaymard, A. (1990). Behavior of 14C aflatoxin M1 during Camembert cheese making. Journal of Environmental Pathology, Toxicology and Oncology, 10, 95e98. Govaris, A., Roussi, V., Koidis, P. A., & Botsoglou, N. A. (2001). Distribution and stability of aflatoxin M1 during processing, ripening and storage of Telemes cheese. Food Additives and Contaminants, 18, 437e443. Horwitz, W. (1982). Evaluation of analytical methods used for regulation of foods and drugs. Analytical Chemistry, 54, 67Ae76A. International Agency for Research on Cancer. (1993). Some Naturally occurring substances: food Items and constituents, heterocyclic aromatic amines and mycotoxins. In. Monographs on the evaluation of carcinogenic risks to humans, Vol. 56 (pp. 489e521). Lyon: IARC. Jouany, J. P., & Diaz, D. E. (2005). Effects of mycotoxins in ruminants. In D. E. Diaz (Ed.), The mycotoxin blue book (pp. 295e321). Nottingham: Nottingham University Press. Kamkar, A., Karim, G., Shojaee Aliabadi, F., & Khaksar, R. (2008). Fate of aflatoxin M1 in Iranian white cheese processing. Food and Chemical Toxicology, 46, 2236e2238.
108
A.M. Fernandes et al. / Food Control 24 (2012) 104e108
López, C., Ramos, L., Ramadan, S., Bulacio, L., & Perez, J. (2001). Distribution of aflatoxin M1 in cheese obtained from milk artificially contaminated. International Journal of Food Microbiology, 64, 211e215. Mayes, L., & McDonald, S. (1995). Aflatoxin M1 in retail milk and milk products. MAFF. CSL Food Science Laboratory. Report No. FD 94/98a. Moss, M. O. (1998). Recent studies of mycotoxins. Journal of Applied Microbioogy e Symposium Supplement, 84, 62Se76S. Neogen, Europe Ltd. (2005). Neocolumn instruction manual. Scotland, U.K. Oliveira, J. S. (1986). Queijo: fundamentos tecnológicos (2a ed). Campinas: UNICAMP. Oliveira, C. A. F., Franco, R. C., Rosim, R. E., & Fernandes, A. M. (2011). Survey of aflatoxin M1 in Minas cheese from the North-east region of São Paulo, Brazil. Food Additives and Contaminants: Part B, 4, 57e60. Oruc, H. H., Cibik, R., Yilmaz, E., & Kalkanli, O. (2006). Distribution and stability of Aflatoxin M1 during processing and ripening of traditional white pickled cheese. Food Additives and Contaminants, 23, 190e195. Oruc, H. H., Cibik, R., Yilmaz, E., & Gunes, E. (2007). Fate of aflatoxin M1 in Kashar cheese. Journal of Food Safety, 27, 82e90.
Park, D. L., & Pohland, A. E. (1986). A rationale for the control of aflatoxin in animal feeds. In P. S. Steyn, & R. Vleggaar (Eds.), Mycotoxins and phycotoxins (pp. 473e482). Amsterdam: Elsevier Applied Science. Pierides, M., El-Nizami, H., Peltonen, K., Salminen, S., & Ahokas, J. (2000). Ability of dairy strains of lactic acid bacteria to bind aflatoxin M1 in a food model. Journal of Food Protection, 63, 645e650. Prado, G., Oliveira, M. S., Pereira, M. L., Abrantes, F. M., Santos, L. G., & Veloso, T. (2000). Aflatoxin M1 in samples of “Minas” cheese commercialized in the city of Belo Horizonte e Minas Gerais/Brazil. Ciência e Tecnologia de Alimentos, 20, 398e400. SAS Institute. (2004). SASÒ user’s guide statistics. Cary, NC: SAS Institute Inc. Scott, P. M. (1990). Natural poisons. In K. Helrich (Ed.), Official methods of analysis of the Association of Official Analytical Chemists (15th ed). (pp. 1184e1213). Arlington: AOAC. Tuinstra, L. G., Roos, A. H., & Van Trijp, J. M. (1993). Liquid chromatographic determination of aflatoxin M1 in milk powder using immunoaffinity columns for clean-up: interlaboratory study. Food Biology Contamination, 76, 1248e1254. Wood, G. E. (1991). Aflatoxin M1. In R. P. Sharma, & D. K. Salunkhe (Eds.), Mycotoxins and phytoalexins (pp. 145e164). Boca Raton: CRC Press.